32:03 Video

How the Scottish Government is Building a Hybrid and Multicloud Strategy with Kubernetes & Portworx

How the Scottish Government is leading the way with their true multi-cloud strategy. Portworx® by Pure Storage is a key solution enabling the portability & flexibility of using any cloud or Kubernetes platform.
Click to View Transcript
Hi, everyone. Thanks very much for joining the session today s Oh, we've got a really interesting session for you. We're going to be looking at a port works user city. Been Scottish government on we're really lucky to have with us today. Neil Smith. Neil, thank you very much for joining the
session. We're going to kick off with quick introductions on. Then I'm going to take you through a bit of an overview of port lax and set the scene a little bit. And then the majority of the remainder of the session will be ah Q and A with Neil when he's going to take us through what he's doing with kubernetes in the Scottish government and also
talk a little bit about how poor works fits into that. And some of the capabilities that attract city and report works in the first place. So before we get into the content, quick introductions, Neil are pass over to you first. Thanks. Thanks, Joe. Hi, everybody. My name is Neil Smith, head of infrastructure.
The Scottish government for one of the largest directorates, which is the agricultural rural economy Directorate. Uncomfortable kind of. The directorate is a as it sounds agricultural base, but we develop applications where farmers concluding for subsidies. We pay out around about a billion dollars and subsidies every year where we develop the
applications. As I said, I manage a multidisciplinary team covering physical, virtual and cloud admits DBS database administrators on multiple Davis platforms. Identity management on multiple doing to management platforms, automation on a bit of cybersecurity, cause and bacon milk. And my name is Joe Gardner on the director of Cloud Native Architecture for the mere region
of pure storage. I work in the Port Moats business unit, and I'm very lucky to work with customers like Neil, who are really on a community's journey, adopting cloud native technologies on anything to do with kubernetes containers and, of course, Port Works falls for then me and my team's remit. So, as I said, we're gonna have a bit of open discussion with Neil about his own communities
journey. But before we do that, I just want to really set the scene a little bit and talk about what I'm seeing brilliant in across different industries when it comes to kubernetes doctrine and actually one of the best sources of information for This is the CNC it surveyed on This is something that they release every year on. If you take a look at this survey that's
publicly available, you'll see that from the data is very clear that Cuban it eases. Now what's gone mainstream? It's cross the chasm. It's now a massive part of pre much every organisations. IittIe strategy is driving lots of transformation in organisations on it is absolutely considered production ready your production suitable by the majority of
communities users. And you could actually see that of those organisations that were hold in the last survey, the vast majority now using community. So it's not just about, you know, those innovative companies were pushing forward on adopting what was previously thought are over cutting edge new technology. We're seeing traditional organisations so large
banks we see manufacturing. We're seeing government such as yourself, Neil now looking at communities a serious platform for for building and deploying their applications. Now Port works also can access surveyed every year on DS from last year survey. There was really interesting question really focused on the maturity of kubernetes and the question was,
which areas do you think need to mature to make it easier to deploy cloud native applications into a community's environment. And as you can see the that well, the top to one of the top responses here is it's storage and data management. I'm into the top two because application development frameworks is another area where therefore works as value and I'll touch on it
very briefly in second. But we can see that storage and data management is absolutely an area that needs to mature from the communities user base on for me. I've been working in this space for a number of years and I've seen that whilst in storage and data management really go up the ranks from the surveys. I mean, if you think back Tiu this sort of five
years ago, it was way down the ranking and I think that's probably because of the kind of applications that were being containerized Then what we're seeing is that as the maturity of communities grows and you saw that 96% of organisations now used in communities figure we're seeing that the applications change. So we're seeing Mawr critical work clothes being moved into kubernetes. On with that comes the need to manage data and
state, and kneel and worship show. We'll hear about that from you in a minute. But for me, that's the reason why storage and data management is becoming an increasingly important area for a lot of organisations because the applications that contain a rising and now becoming more critical is those line of business banking acts, for example, or or use the data basis.
So you'll hear a lot from from some of my colleagues around the port Works report works platform, but I just want todo quickly give you sort of ah, overview a little bit of context. So you've got got in mind all the different pieces before we go in to a specific use case on, Do you hear us talk about port Works being a platform?
And the reason we do that is because it spans lots of different areas I like to think of. It is kind of being talked down from another software on, apart from the service, respected and sort of bottom up from an infrastructure perspective on what I mean by that is that there are software, the service components to poor works which touch on that application framework point you'll hear of our poor workspaces services.
Just manage databases of service platform You hear about poor work back up as a service and AWS. But then underpinning That is a foundation of data management capabilities around the high availability in a single cluster disaster recovery for multi cluster security and encryption tools. Support ability, which I'll go into in more detail on a moment.
And really, those foundational capabilities really important to platform Milner's. But they're the ones who are taking. They're kubernetes distribution of choice. And, of course, Port Works reports, all of them on building these as a service platforms for their developers to consume on. Really, the ultimate objective here is to
support any application and production. And that's really the takeaway in terms of the value of four works is that by combining these as a service of their components with these industry leading data management and infrastructure components, platform builders can provide an environment where any state full application could be run. Well, what I want to dio is just going to a little bit more detail on a specific use case
that comes up a lot on, You know, I've mentioned for a moment, this idea of data possibility on his side port works. There's a component called PX Migrate on dis is really a very simple tool that takes advantage of the agnosticism. Support works brings to your communities environment for managing application
data, and it uses the same kind of design principles that true in communities and containers themselves to make. The report works volume in the application data as portable as the container. They actually heard one of our customers referred to a port. What's volume as a little bit like a data container?
Now the beauty of that is that it can then be moved free between any environment on that means different crew. But that's his distributions. Different infrastructure environments on famous Data Sentenced to Public Cloud You really have the freedom as a user of this capability to make you to make your days are truly possible on. But I think the key thing to bear in mind is
that without the application being part of that, that portability that migration. It's very difficult to maintain any kind of consistency, and I think that's where Port works is unique, its ability to not only my great data, but also take all of that application in conflict with it. So from the use of perspective kind of single click on entire applications
that will pop up consistently in any other Cuban exiles distribution running anywhere else in the world. So this is a really powerful capability. And, as you can probably imagine, a real driver behind poor works adoption. It actually leads us nicely into into a bit of a Q and A with Neil because when near than I first met, when we first talked about port back. So I think this was probably probably the area
that let's douse that you if I remember correctly, I mean kneel before we get into the detail of the of the migration capability. It would be good just to maybe get a little bit of background on your communities. Use that the Scottish government and kind of the journey you've been on with the technology? Sure. Thanks. Thanks you.
Hey, everybody. So we're very much still in their infancy. I say that all the time in terms of even on our cloud journey. You know that very much. Everyday's day, One etcetera s. So we're we're still evolving and elaborating into our in to kubernetes journey. Always evolving, we're looking at migrating away from Dr Containers and £2 man
containers and looking at security angle, etcetera. But way we know it's the real. The real value wasthe was with the number of containers we were running. We need started todo to orchestrate them on day. Obviously, communities was the only logical choice. And there's many different flavours in terms of an enterprise kubernetes platform.
Andre. You'll probably find out through the season questions that Joel ask on. You'll find out that I place a large emphasis on trying to be agnostic as possible. You know, you always get tied in from some point in place, but trying to be agnostic as possible and and the orchestration of containers is no different. And I think over the series of,
however excellent many questions Joe was gonna ask you. We're going to find out how we we took our kubernetes implementation and how we took it to the which I would classify as the next level and to make it as agnostic as possible, but then puts their control and power back into from a customer standpoint. My hands Yeah, and I think we often focus on the
technology more than actually some of the outcomes of of using that technology. So be good just to kind of get a picture from you off, I guess, more the business side of things, right? What kind of underpinned? The adoption of containers in the first place? Five Scottish government. I mean, what was always kind of the old
objective behind using this new technology, I think with E, we're no different all over government organisation. We're no different to any enterprise company out there, etcetera. Everyone's over the last few years you hear about you've got to be more day evolves, more agile, etcetera in that Michael service
approach, right? You want to be more micro services, etcetera, and we're no different on D. That's where container ization comes into its into its own. Because you can s so far easier adoption into that Michael service support, etcetera. Until small components what is previously in
the government, You know, we were we were pretty successful that building monolithic applications s o if you build up more of a micro serviced approach, etcetera. But from the business angle on that is ah, more make service approach so you can give a lot more iterative approach to the business, eh? So you can find out, ran and doing that
waterfall method do that more agile can approach. But it's the path to live and the and the velocity of live from taking something that's on idea to actually delivering something that business confined, benefit worth. You know, like that MVP Kenny Piece on, then build on it from there. But it's traditionally we were very waterfall
method where we get the requirement will spend, I don't know, a month, six months a year. You get to the end and in the business gonna go well, that's not what we wanted. Do you like Oh, well, that's kind of what you've got. So the so the kubernetes approached further emphasised the need to beam or a micro services. So our service,
or or in taking architecture approach and more agile on then the business can then actually particularly have an interest in governmental weaken. There's a lot of new schemes and regulations that get introduced very attractively and sometimes out of nowhere, it gives us the ability to be t react very quickly. and deliver very quickly. So it was so that business could see a huge
benefit in it because of the that the rate of velocity we can take it from from David to probe. Yeah, I guess that's where the kind of the public Kyle piece comes into the picture as well, right? Because I know that there's quite a lot of public power use, for I was quite a lot or beginning to be a lot of public clarity's that Scottish government.
Yes, so there was. There's there's numerous different directorates inside the Scottish government, and also we're no different. We do centralize up some of those functions. A lot of them are independent, and we were no different from that. So, like, say, we're we're we're always evolving. We're always Kanae moving forward,
etcetera in terms of how are our cloud use but close here to stay sane with kubernetes and you've got to acknowledge that Go on in the days now where, you know, back in my early days, it was all hardware based on sister administration, etcetera. You know the role of a sister administrators going now, right? And the those days were phasing out.
We look at in Serverless, for example, the and I mentioned maybe I should maybe qualify that, but I mentioned about being agnostic as possible. Then I mentioned a serverless king. The solution is the way I always put it. Joys as you should always host your workloads. Based on the requirements I find a A, particularly the government. You find the Kenny mantra of its cloud First
cloud first. The cloud first doesn't mean everything should go in the clouds. But that seems to be what happened, says and right now is you get someone that has Ah, a new application says we will stick in the cloud but it might not be suited to the clouds. Similarly, with our on legacy transformation project all stick it in the cloud.
It's not always the case and the way I tried to break that down, trying to come up with an analogy that made made senior civil servants get their head around. It wasthe because I've seen numerous government departments just stick everything in the cloud, and it was more I asked, baste. It's like, well, you know, you're really just
running a day care centre in the cloud, and that's a sure fire way to spend a lot of money. Thenardier. I put Round two. Seen some servants. Woz. It's a bit like if you go to the doctor and that doctor prescribes you the same prescription that he or she has prescribed everybody else in the day.
That doesn't happen because everyone's unique and it's the same with a new application. It's the same with them. Legacy Transformation Pieces is a look at his requirements and hosted when it's applicable, whether that's on a physical platform, probably highly likely in a virtual platform, maybe on a container on a virtual platform, maybe on a container in the cloud, perhaps cloud native service,
perhaps on a serverless solution. You've all these options, so So actually, hostiles were close based on those requirements on. It's just getting into that that that may say, And so we go back to Maryland Point about what you're talking about, serverless. And you're talking about Agnes Agnes necessary in in terms of the comments I made earlier were service really ties you and you can't lift and
shift that you know, from from, say, one hype scale provided to another. But when you base it on its workload when you based on its requirements, it fit fitted. The bill effectively gave the business what they wanted. It cost us $8 a day to run. And operationally, those day two overheads are no longer there.
So I mentioned by the traditional society minutes so forthright. But they no longer have to maintain that physical system or with the virtual system. It's all it's all maintained there, and that was the right for that solution. That was the right thing to do. Everybody one on that but on depend on how your questions go later on Joe's. I'm very big on the culture aspect of the team
is I want to give my guys the time and space to be able to be creative and innovative. It's possible. So happy persons a productive person. So if we can implement a serverless solution, actually, maybe minimizes those operational overheads that would normally traditionally have for all normal applications, then I so that gives him more free time to be able to experiment and play.
And I'm a very big believer in my role as head of infrastructure. Is is is to be an enabler for these guys to be creative as possible and shield them from maybe some of that bureaucracy and red tape. Let them be as creative as possible. You know, there's still got to do the day to day rolls and pass etcetera in that gap.
Let them be, you know, exploring and creative. And that's part of the engagement that we came with with with Port. Works were pure storage customer. We were the first in Scotland to adopt pure storage. I think itwas 2015. So you know, we have been long term customer, but again it was It wasn't a risk. It was.
It was. We took a proof of concept. We ran the box. The pure storage array on bond more than more than smashed our requirements on, Ironically, Ondas very difficult. The government. It turned on a narrow I. It was going to get a detail, but not for this. This conversation and it's not going to have a
narrow I inside the government. But that solution did. But I have said is one of the guys had mentioned about hey, pure storage of just about port works and we're doing this community's peace and looks like that will manage the data layer because, you know, although, you know, containers by their very nature, you always think about they're gonna be more kind
of stateless, right? And you know, of any stateless applications, like they're also the surgical applications. And that was where it was born from. On them engaged with you, Joe. And then yeah. Then we we don't run a pretty concept to go from there.
Sorry when I've been a little bit of a tangent job now and also No, I mean, it's great to get the context. I think you just mentioned for a couple of themes that kind of touched one of the start around agnosticism. And yet the fact that actually, the majority of applications aren't status even though, absolutely, you know, five years ago or probably further back,
he was almost a religious argument around whether container should be status stable. But the reality is they're actually on. We'll go out to you move in there from let's call it a traditional organisation. The wall ST comes into the picture, right? Yeah, but I want to go back to that point on agnosticism because, you know, we've both mentioned it quite a few times.
It is the sign of a kind of theme of the conservation. So maybe maybe be good to hear about. We've thought up having a cloud talked about. You know how you're using kubernetes to kind of make make your team's morale. Joe, I'll give you your own teams the I guess, the room to breathe to be more innovative. How do you you see kind of port works fitting into that?
And also, I guess, you know, we have that specific use case around portability. Be good to be good to hear about that and how that fits and perhaps a longer term vision as well. Sure, So a big question questions in one. No, no, it's cool. It's cool. So I get I enjoying doing these type of
sessions. I get around a bit like being on stage and talking to and from that I get to engage with a lot of guys like yourself, Joe and and and your peers and the, you know, the the sea level sweets, etcetera, and I manage to get conversations with them, and it was very keen on multi class for me. Multi cloud is going to be the future on.
I would ask a lot of guys like yourself do on C, T or C I ose and went here you do multi cloud and they would say, Yeah, we are great. How you doing it? They would say, Well, we've We've got work Cody and eight of us and work will be in Missouri and I'm like, OK, and how he says a few to switch that around and they would say, Oh, that's so that's that we can't do that take a while long.
Would that take months, if not years like Well, that's a That's a bit of a risk. And I always present the questions is But what happens there, Father? All ah, security vulnerability was found on a hyper scale cloud. Provide a platform that they couldnt remediate quickly. All they put their costs up overnight.
How quickly you move off that and you're right, you know, that's very difficult. Well, that's not true. Multi clouding. True multi cloud is by be able to run your workloads on any anywhere you want effective than any high school cloud provider. Yeah, that's where kind of port works. Can it came in right?
Because if you're going to be able to run those workloads anywhere you want, Then you're gonna have to manage the data later, right? Because data is key in this and the only way to do true multi cloud is continue ization and kubernetes. There's no we get around that. But you've seen your stat the earlier on their job out Think it was.
95% of people have adopted or are currently bringing proving concepts and kubernetes is the future. So using true multi cloud and having to run your workloads wherever you want. Whenever you want this hugely powerful how do you manage that data later on? That's when we engaged with yourself Joe on Deport Works Team. Where we decided to come up with was Who sent
you guys a challenge with yourself and my guys on the team to saying, Okay, we've got our we've got in on premise whole host of on premise workloads. But let's use this workload. A running inside of'em were platform on on on premise data centre inside open shift on, but we want to do is is we want to take that workload is traditional. Just have front and back end application.
Good day to beat nothing. Special job right with them? I think so. What if they want to do is over complicated with some of the crazy acts that we have, right? It was a pretty concept to begin with, but yes, Java based with my secret backing on but I wanted to do was take that workload that's on premises.
Have you ever stay inside open shift and running into eight of us on DWI had stood up a Rosa instance The red hot open shifts service on AWS. But it could have easily have been ah, bunch of the C two instances, etcetera. We just done the manage service calls a proof of concept. It was easier to spend on Central on Bond.
What We've been done with yourself, Joe on the guys as we managed to migrate out that that application of discussed and run into into AWS on the thing I was blown away with was it took It took a minute and a half and I was like, All right, well, actually, you know, it took me back to merely of'em were days when we were doing kind of emotion and migrate and virtual machines from host to host etcetera.
Like what? Just this is this is really cool technology. But then I got greedy, and I says, Well, why don't we migrate over into and in tears your you know, into open shift and on Azure. So migrated that over again. A man and a half. And then we migrated it back down from Missouri back into the on premise bmo Estate on open shit.
So that that was hugely powerful, right? I was like, Wow. Okay, so now we can have the ability to move our work lower containerized workload than a cabernets platform to any high school cloud provider. But then again, you got greedy, and I was pushing you join men? What else can we do?
I know it does. X y zed in the PX back up and stuff. And we were just proving the PX, migrate and Ph. D Ark any component, and then you can you mentioned you're like, Oh, well, you know, we've got the ability to to migrate away from your open shift a platform. And I was like, Okay, so that sounds really
exciting. So we so we way continued the proof of concept where we then took that on premise workload on the M. We're on open shift, then ran into AWS on tickets and then take that same workload and then running into a zero on a K s and then back down onto a picture. So for me, I was going to blown away with that because that gave us the capability to know on
Lee run our workloads on any HEIP scale cloud provider, but no longer be tied to any enterprise kubernetes platform. Now you're gonna be We're gonna be clearly aware you called out early, Joe. You were talking about when you use an open shift. If you're using roots, then you know they're native to red hat and open shifts when you're migrating over.
And you mentioned about taking that application state and your your example slide for the earlier on is you have to be acutely aware that if you use the roots that you're gonna have to remedy that in the migration of the new platform. Similarly, with them, a rancher uses projects etcetera. So there's those little video sync. Chrissy is right,
but they're not there. No, insurmountable are difficult to overcome or you can treat so true close to the kubernetes natively is possible. That makes that kind of migration possible. But for me, if she's a powerful, I could run my workloads anywhere I want. Any time I choose relatively quickly, but also as well that I'm not tied to that Enterprise
Cooper next platform and from a customer point of view, that's powerful because you know, my red hat renewal becoming up, etcetera, we would be about a writ, large rate hat state running from rail, TJ, boss, middleware and all whole host of other stuff. We have open shit and they might put the prices up.
You know, right now the point of us off of talking Joe, the pound against the dollar isn't great. And generally, when we buy a software against big, large American companies, it's always based against the dollar. So it's naturally going to go up and praise. But yeah, for me, I've got the ability now to say, Actually,
no, I'm not going to pay that price. I'm in a much more powerful negotiating piece to say, because I can just click a button and actually, my great my work clothes on an open ship platform to any other companies base platform. So I haven't run that proof of concept. That was a kind of a real game changer for us to go. That's doing true.
Multi cloud on is so agnostic because we're no, we're not tying ourselves to take any anything or anywhere on bond particular from a government organisation coming at it from that angle, which could be quite unique, right? Governments traditionally are seen at bottom end of the curve. Where is you know, as I will say it is particularly and my team is, is we're always at the top end of the curve,
but it's it's all about that creativity and so forth. And so it was a hugely powerful proof of concept that, you know, that we can then actually say Okay, no, only have we thought about doing it, we've actually proven it on. We're actually beginning to enterprise out because, you know, there's other things we have to manage on a multi cloud kind of world.
Whether that's network and etcetera, there's three main components I'll say is you've got the detail we arrive in Port Works, matches that box. We've also got the network and later. How you imagine that? And also security compliance. And what would that is over the the time we were engaged with you,
Joyce, as we've started to enterprise out. So not only did we come up with a great idea worked with you guys smash the terrific concept, particularly date earlier. We've also no proven the multi cloud network and peace and good and compliance. So now we can actually see. Here's a great idea. It's actually been proven on. It's been enterprised up and they will roll out.
Except it's a real powerful storey for something that was relatively quick and easy to implement. I know we will have the ability for anyone to ask questions, but we managed todo all that you know, and inside a small amount of time with the small amount of people it was Yeah, it was Ah, yeah, my mind was blown again to be battling said earlier of'em were days And be emotional like
this is really cool. I guess you just get back to you. I'm glad to hear that mother away. It was It was It was great fun. It was agreed to give us the right, but just going back to your point on, I guess. The vendor side. I mean, the data is a waste of sticky bit right, and it's a topic for another day,
and we should probably talk about it as well. But again, that's where PDS comes, then into the picture, being giving you that database of service like experience that completely agnostic, bring around environment. Right? So we were only about a minute left, but just to wrap up,
I guess it unopened question for you kind of lost in the future for you, right? I mean, up your import with uses love to hear about these kind of ongoing journeys and hear about what one other users and customers are doing. So you know what's on your road map for your your your task list? I guess if you look ahead over the next 12
months, yeah, za son, you mentioned. And so the PDS pieces is really of interest, innocence where we're trying to strike towards as it is, we're trying to do everything. Infrastructures called security is called trying to automate as much as possible, but then also provide a self service like said is we have our we have a large number of
developers we have it fluctuates 102 150 developers, etcetera. Certain point. Sometimes we want to give the self service, and the database has always been a challenge for us. Ironically, with pure storage, we you snapshot cloning with answerable to give our developers a copy of the Prod database and on a miser, the click of a button with a footprint.
When I heard about the PDS service with port Works, where we could get a developer is just too self service themselves to pick any any any day. We have to choose on any version and then taken. Just select it, spin it out. What we need to do to take it back down is that very much fits into into my my larger strategy where ideally,
it's all that software defined data centre where infrastructures called doing security is going. That's another point for another not talk for another time, but off to me as much as possible. Give self service, which again goes back to my earlier point, is is, the more we can automate more incorporates of their self service frees up mawr time for for my guys to do cool new stuff
on your super excited about again and then get opportunities like here. Where is it? You know, I'm just a guy that just talked about It's all those guys at the genius, but they will enjoy what they're doing. And I really enjoyed doing this kind of stuff. Buys. It's a win win.
Yeah, I know the feeling. By the way, the deal thank you were at the time, but thanks again for joining the session and being so open about your you know, your journey with communities and port works in the area. Directorate On thanks to the audience for joining the session today. We hope you enjoyed it on that. You know, it was a useful,
insightful session point. Thanks very much.
  • Government
  • Portworx
  • Video
  • Pure//Accelerate
  • Multicloud

Join Neill Smith, Head of IT Infrastructure, Scottish Government, and Joe Gardiner, Head of Cloud Native Architecture, Pure Storage, who discuss how the Scottish Government is leading the way with their true multi-cloud strategy. Portworx® by Pure Storage is a key solution enabling the portability & flexibility of using any cloud or Kubernetes platform.

View All Pure//Accelerate Sessions On Demand

Watch Now
Continue Watching
We hope you found this preview valuable. To continue watching this video please provide your information below.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Meet with an Expert

Let’s talk. Book a 1:1 meeting with one of our experts to discuss your specific needs.

Questions, Comments?

Have a question or comment about Pure products or certifications?  We’re here to help.

Schedule a Demo

Schedule a live demo and see for yourself how Pure can help transform your data into powerful outcomes. 

Call us: 833-371-7873



Pure Storage HQ

650 Castro St #400

Mountain View, CA 94041

800-379-7873 (general info)


Your Browser Is No Longer Supported!

Older browsers often represent security risks. In order to deliver the best possible experience when using our site, please update to any of these latest browsers.